Molecular genetic investigations on Austria's patron saint Leopold III.

The successful marriage policy of margrave Leopold III increased the importance of the House of Babenberg in late medieval Austria (12th century). Historical documentation is inconclusive in providing evidence whether or not his eldest son Adalbert derived from an earlier relationship or from the marriage with King Henry IV's daughter Agnes of Waiblingen, with whom Leopold is considered to have had 17 children. As a matter of fact Adalbert was ignored in the line of succession in favor of a younger brother, Leopold IV, which has led to long term historical discussions. Human remains attributed to these individuals were subjected to DNA analysis. Autosomal, Y-chromosomal and mitochondrial DNA analyses brought successful results, which suggested that Leopold III, Agnes and Adalbert were related in parent-son constellation, in contrast to historical considerations. A possible mix-up of Adalbert's remains with those of his younger brother Ernst could not be confirmed by DNA analysis.

Introduction

Leopold III, the Saint, was the Austrian margrave from 1095/1096 until his death in 1136. His successful marriage policy led to substantial economical and political wealth of the House of Babenberg, which ruled Austria between 976 and 1246. With his second wife Agnes of Waiblingen, daughter of the Salian King Henry IV, he was reported of having 17 children. Whether or not a possible earlier marriage resulted in offspring has been a matter of debate among historians. Although the source material situation is unclear, his first son Adalbert is generally believed to originate from an earlier relationship. A strong indicator for this assumption is the fact that Adalbert was ignored in the line of succession in favor of Agnes’ son, Leopold IV, who died in 1141. Leopold III was canonized by pope Innocence VIII in 1485. His remains were buried in the middle one of three tomb chambers in the monastery of Klosterneuburg, Lower Austria, which he had founded in 1114 [1]. In the northern chamber, Adalbert's remains were laid to rest, when he died around 1138, shortly after his father. The southern chamber was assigned to Agnes who died in 1143. The aim of this study was to investigate whether Adalbert was the common son of Leopold III and Agnes or derived from an earlier, potentially unknown, relationship of Leopold III.

Materials and methods

The putative skeletal remains (femoral bones) of Leopold III (L), Agnes (AG) and Adalbert (AD) from the monastery of Klosterneuburg were delivered to the Institute of Legal Medicine, Innsbruck Medical University. For a second set of experiments we later also received a femoral bone from the monastery of Heiligenkreuz, attributed to Ernst (E), son of Leopold III and Agnes, who died in 1137. Small pieces (ca. 2 cm × 1 cm × 1 cm) were excised from each specimen with a bone saw and subjected to mechanical surface cleaning with sterile scalpel blades. Samples were bathed in sodium hypochlorite (≥4% active chlorine) at room temperature for 15 min, spoiled in purified water (DNA/RNA free), rinsed in absolute ethanol for 5 min and exposed to UV light for 10 min. Samples were dried in a closed laminar flow cabinet over night and then powdered using a vibrating ball mill (Laarmann Group BV, Roermond, The Netherlands). Lysis and DNA extraction were performed according to [2]. Buccal swabs were taken from individuals involved in sample handling and extracted using Chelex (Bio-Rad Laboratories, Hercules, CA, USA) according to [3] to detect any potential contamination.

The quantity of mitochondrial (mt) and nuclear (n)DNA from the bone extracts was determined using a real-time PCR approach according to [4,5] on a 7500 Fast Real-Time PCR System (Life Technologies (LT), Carlsbad, CA, USA). Autosomal Short Tandem Repeat (STR) profiles were established using the AmpFℓSTR NGM Select PCR Amplification Kit (LT) and the Powerplex ESX and ESI 17 Systems (Promega, Madison, WI, USA). The Familias software package (http://familias.name/) [6] and DNAVIEW software (version 29.68) designed by Charles Brenner (Berkeley, CA) were used for kinship likelihood calculations based on STR allele frequencies derived from [7-10]. Y-STR analyses were performed using the AmpFℓSTR Yfiler PCR Amplification Kit (LT) and the YHRD database (http://www.yhrd.org/, release 39 [11]) was used to query the resulting Y-STR haplotypes. MtDNA analyses were performed using a midi-amplicon approach described in [12]. Sequences were aligned and interpreted according to [13] using the Sequencher software (V. 5, Gene Codes Corporation, Ann Arbor, MI, USA). MtDNA control region haplotypes were queried in the EMPOP database (http://empop.org/, release 7 [14]).

Results

The femoral bones attributed to Leopold III (L), Agnes (AG) and Adalbert (AD) were very well preserved and showed dense compact bone structure (Fig. S1). The mechanical and chemical processing of the samples was performed with the appropriate care required for challenging samples [15,16] and the derived DNA quantification values were plausible with respect to the sample conditions (Table 1). Extraction blank quantification values did not indicate relevant DNA contamination and STR as well as mtDNA typing did not yield interpretable DNA profiles. All results obtained in this study were different from the DNA profiles of all laboratory staff involved in the analyses (internal elimination database) and individuals who handled the samples outside the laboratory (data not shown).

Table 1

DNA quantities, autosomal STR profiles, Y-chromosomal STR and mtDNA haplotypes of long bone samples attributed to Leopold III (L), Agnes of Waiblingen (AG), Ernst (E) and Adalbert (AD). b.d., below detection limit; nps, nucleotide positions with respect to the rCRS [17].

Sample-ID	Amelo	D1S1656	D2S441	D2S1338	D3S1358	FGA	D8S1179	D10S1248	TH01	vWA	D12S391	D16S539	D18S51	D19S433	D21S11	D22S1045
Autosomal STR loci
AG	X, X	15, 16.3	10, 11	18, 18	18, 18	26, 26	9, 14	14, 16	6, 8	17, 18	18, 21	11, 12	12, 14	14, 14.2	28, 30	13, 16
AD	X, Y	16, 16.3	10, 10	18, 23	16, 18	21, 26	9, 12	13, 16	8, 8	17, 17	18, 18	11, 12	14, 15	14, 14	30, 30	15, 16
E	X, Y	–	14, 14	(23), (23)	(15), 18	–	–	13, 16	8, 9.3	18, 18	(16), 21	9, 12	(14), (15)	(12), (12)	28, 28	16, 16
L	X, Y	15, 16	10, 14	19, 23	16, 18	20, 21	12, 12	12, 13	8, 9.3	17, 18	18, 18	11, 14	15, 17	14, 16.2	30, 30	15, 16

The sex-specific DNA test (Amelogenin) produced male genotypes (Amelo X, Y) for samples L and AD and a female genotype (Amelo X) for sample AG (Table 1), which was in agreement with what had been expected. Those three femoral bones yielded full autosomal STR profiles with the exception of SE33 that gave no results most likely due to its relatively large fragment size. The successfully analyzed 15 STR loci revealed allele sharing patterns of sample AD with samples L and AG as expected for a son-parent constellation, i.e., one allele shared per locus with each parent (Table 1). We compared the three hypotheses of the donor of sample AD being the son of both the donors of sample L and AG (H1), the son of L (H1a) and the son of AG (H1b) to the alternative scenario that the profiles derive from individuals completely unrelated to each other (H0), using Familias. The corresponding likelihood ratio (LR) brought values of 1 × 105 in favor of H1a, 2 × 107 in favor of H1b and 1.32 × 1014 of H1 over H0. This provides strong support for the hypothesis that the donor of sample AD is a son to L and AG. These findings were also confirmed by DNA-View resulting in LRs of 7.5 × 104 (H1a), 1.6 × 107 (H1b) and 5.5 × 1013 (H1) each supporting paternity. The results from both uniparental marker systems corroborate those from autosomal DNA analyses: Y-chromosomal STR typing revealed an identical haplotype between AD and L in the overlapping successfully analyzed loci (with the exception of an additionally observed allele 10 in DYS391 in sample AD; Table 1). The haplotype was found 23 times in 13,143 European males (YHRD, release 39). The analysis of the mtDNA control region yielded identical haplotypes (barring length variation at the highly mutable homopolymer cytosine tract around nucleotide position 309) between AD and AG that belong to haplogroup R0 (Table 1). Disregarding length variation in C-tracts, this haplotype was not observed in 4630 west Eurasians (EMPOP, release 7).

The compact bone structure of the bone sample E from Heiligenkreuz, attributed to Ernst, was lighter and less well preserved than the remains from Klosterneuburg, which was reflected by much smaller amounts of extracted nuclear and mtDNA (Table 1). The mini-amplicon assay for mtDNA typing gave a full control region haplotype (haplogroup T2b [12]) that was different to the one obtained from AD and AG. Autosomal STR typing (34 cycles, 2 independent amplifications, compound interpretation method) gave a partial STR profile with exclusions in two STR loci relative to the genotype of sample L (Table 1). Under the assumption of these two mutation events DNA-View brought a likelihood ratio of 4 × 10−4 for the hypotheses of the donor of sample E being the son of the donor of sample L (H1a) to the alternative scenario that the profiles derive from individuals completely unrelated to each other (H0).

Discussion

The samples from Klosterneuburg were investigated under the assumption that they constitute human remains of margrave Leopold III, his wife Agnes of Waiblingen, and his son Adalbert, bearing in mind that such old remains might have been subject to replacement or exchange during history. DNA typing was used to test whether or not the putative remains of these three individuals constitute a full biological family or only indicate father–son relatedness, as expected from historical sources. Successful DNA typing results provided strong support for relatedness in a parent–son constellation for the samples L, AG and AD, opposing the historical expectation. Consequently, an alternative scenario came under consideration: Adalbert could still be the son of an earlier wife of Leopold if the tested remains (AD) were actually from a different male descendant of Leopold and Agnes. There is historical evidence supporting this assumption that was brought to our attention after completion of the first set of experiments: in the middle of the 13th century, substantial renovations took place in the monastery of Klosterneuburg. It was possibly on behalf of Friedrich II, the last of the Babenberg Dynasty, that before 1240 the remains of his ancestors were transferred to Heiligenkreuz, a monastery also founded by Leopold III in 1133. Not only Adalbert, but also his brother Ernst, who died in 1137, were to be buried there together, the tombstone still bears both their names today. However, when the putative double burial was opened in Heiligenkreuz in 1739, skeletal remains of only one individual were found, carefully arranged in a transport box. The set of remains assigned to AD were also found in a transport box in the northern tomb chamber of Klosterneuburg in 1979. This leads to the suspicion that both boxes might have been transferred from Klosterneuburg to Heiligenkreuz in a first attempt, but due to lack in space – the tomb in Heiligenkreuz was designed in small dimensions, maybe too small for both burials – only one transport box was actually buried there while the other one was returned to Klosterneuburg. Alternatively, both skeletal remains might have been prepared for transportation but only one of them was actually transferred to Heiligenkreuz in appreciation of the small dimensions of the tomb there. One way or the other, both scenarios leave room for speculation that the transport boxes might have been mistakenly exchanged and thus the remains of Ernst could have ended up in Klosterneuburg instead of those of his brother Adalbert.

To test this hypothesis we investigated a femoral bone sample attributed to Ernst (E) from Heiligenkreuz that was sent to our laboratory for a second set of analyses. Autosomal STR typing provided evidence against close relatedness of samples E with L and AG, suggesting that those remains belong to a different, unrelated individual to Leopold and Agnes.

Conclusions

The DNA analyses on bone samples attributed to Adalbert, Agnes and Leopold III gave strong evidence of a son-parent constellation, a result that is discordant with the expectations based on historical considerations. A possible exchange of samples between Adalbert and a younger brother, Ernst, son of Leopold III and Agnes according to historical sources, could not be confirmed as DNA analysis results indicate an exclusion of fatherhood of Leopold III for sample E. Based on the samples provided for DNA typing, it seems most likely that Adalbert was indeed a son of Leopold III and Agnes of Waiblingen.